TechMiner: Extracting Technologies from Academic Publications

Abstract

Abstract. In recent years we have seen the emergence of a variety of scholarly datasets. Typically these capture 'standard' scholarly entities and their connections, such as authors, affiliations, venues, publications, citations, and others. However, as the repositories grow and the technology improves, researchers are adding new entities to these repositories to develop a richer model of the scholarly domain. In this paper, we introduce TechMiner, a new approach, which combines NLP, machine learning and semantic technologies, for mining technologies from research publications and generating an OWL ontology describing their relationships with other research entities. The resulting knowledge base can support a number of tasks, such as: richer semantic search, which can exploit the technology dimension to support better retrieval of publications; richer expert search; monitoring the emergence and impact of new technologies, both within and across scientific fields; studying the scholarly dynamics associated with the emergence of new technologies; and others. TechMiner was evaluated on a manually annotated gold standard and the results indicate that it significantly outperforms alternative NLP approaches and that its semantic features improve performance significantly with respect to both recall and precision.

Forecasting the Spreading of Technologies in Research Communities

“…In general it is possible to generate a set of technologies by i) extracting them from research papers by means of automatic methods [12,13], ii) obtaining them from a manually curated software repository (e.g., the Resource Identification Initiative portal [14]), or iii) getting them from a general knowledge base (e.g., DBpedia [15]). Since the focus of this study is on the analysis of technologies and not on their identification, we created a manually curated set of technologies in the fields of Semantic Web and Artificial Intelligence.…”

“…We first selected an initial set of about 2,000 technologies by running TechMiner [12] on a set of 3,000 papers in Semantic Web in order to find technologies that were originated or adopted by this field. We then manually cleaned and enriched the resulting dataset by discarding incorrect results and we also added 500 other technologies sourced from Wikipedia pages listing Artificial Intelligence and Machine Learning algorithms and methods.…”

“…In our implementation of TTF, the sequence of matrices is materialised as a json file that is fed to the forecasting module, which will use it for predicting which technology will emerge in which topic in the years to follow 12 .…”

“…12 The technology-topic matrices and the TTF code are available at http://rexplore.kmi.open.ac.uk/TTF/downloads/code_and_data.zip.…”

TSE-NER: An Iterative Approach for Long-Tail Entity Extraction in Scientific Publications

“…Literature [20,26,27] shows that training of domain-specific NER/NETs is still an open challenge for two main reasons: (1) the long-tail nature of such entity types, both in existing knowledge bases and in the targeted document collections [22]; and (2) the high cost associated with the creation of hand-crafted rules, or human-labeled training datasets for supervised machine learning techniques. Few approaches addressed these problems by relying on bootstrapping [27] or Entity Expansion [3,11] techniques, achieving promising performance.…”